• International Journal of Computer Science & Network Security
• /
• v.23 no.10
• /
• pp.115-128
• /
• 2023

Deep learning has been incorporating various optimization techniques motivated by new pragmatic optimizing algorithm advancements and their usage has a central role in Machine learning. In recent past, new avatars of various optimizers are being put into practice and their suitability and applicability has been reported on various domains. The resurgence of novelty starts from Stochastic Gradient Descent to convex and non-convex and derivative-free approaches. In the contemporary of these horizons of optimizers, choosing a best-fit or appropriate optimizer is an important consideration in deep learning theme as these working-horse engines determines the final performance predicted by the model. Moreover with increasing number of deep layers tantamount higher complexity with hyper-parameter tuning and consequently need to delve for a befitting optimizer. We empirically examine most popular and widely used optimizers on various data sets and networks-like MNIST and GAN plus others. The pragmatic comparison focuses on their similarities, differences and possibilities of their suitability for a given application. Additionally, the recent optimizer variants are highlighted with their subtlety. The article emphasizes on their critical role and pinpoints buttress options while choosing among them.

• ETRI Journal
• /
• v.45 no.5
• /
• pp.822-835
• /
• 2023

Recent semantic segmentation frameworks usually combine low-level and high-level context information to achieve improved performance. In addition, postlevel context information is also considered. In this study, we present a Context ReFinement Network (CRFNet) and its training method to improve the semantic predictions of segmentation models of the encoder-decoder structure. Our study is based on postprocessing, which directly considers the relationship between spatially neighboring pixels of a label map, such as Markov and conditional random fields. CRFNet comprises two modules: a refiner and a combiner that, respectively, refine the context information from the output features of the conventional semantic segmentation network model and combine the refined features with the intermediate features from the decoding process of the segmentation model to produce the final output. To train CRFNet to refine the semantic predictions more accurately, we proposed a sequential training scheme. Using various backbone networks (ENet, ERFNet, and HyperSeg), we extensively evaluated our model on three large-scale, real-world datasets to demonstrate the effectiveness of our approach.

• International Journal of Computer Science & Network Security
• /
• v.24 no.2
• /
• pp.196-202
• /
• 2024

Currently, the second most devastating form of cancer in people, particularly in women, is Breast Cancer (BC). In the healthcare industry, Machine Learning (ML) is commonly employed in fatal disease prediction. Due to breast cancer's favorable prognosis at an early stage, a model is created to utilize the Dataset on Wisconsin Diagnostic Breast Cancer (WDBC). Conversely, this model's overarching axiom is to compare the effectiveness of five well-known ML classifiers, including Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbor (KNN), and Naive Bayes (NB) with the conventional method. To counterbalance the effect with conventional methods, the overarching tactic we utilized was hyperparameter tuning utilizing the grid search method, which improved accuracy, secondary precision, third recall, and finally the F1 score. In this study hyperparameter tuning model, the rate of accuracy increased from 94.15% to 98.83% whereas the accuracy of the conventional method increased from 93.56% to 97.08%. According to this investigation, KNN outperformed all other classifiers in terms of accuracy, achieving a score of 98.83%. In conclusion, our study shows that KNN works well with the hyper-tuning method. These analyses show that this study prediction approach is useful in prognosticating women with breast cancer with a viable performance and more accurate findings when compared to the conventional approach.

• Journal of the Korean Society of Industry Convergence
• /
• v.27 no.4_2
• /
• pp.897-908
• /
• 2024

Evaluating the adequacy of facility safety inspection and diagnosis services performed by private enterprises is a time-consuming and administratively complex process. This study aims to analyze the determinants that could influence the rating of these safety inspection and diagnosis services using data analytics approach. Through a comparative analysis of several machine learning algorithms suitable for multi-class classification, we selected the model with the best performance (Random Forest) and identified the main determinants using the permutation importance technique. Among the variables examined, "contract value," "days of service performed" and "adherence to fair market value" were found to be strongly correlated with the rating assessments. Furthermore, we discovered that the skills and expertise of service performing personnel significantly impacted the rating. The results of this study can contribute to the enhancement of the current post-evaluation administrative processes and offer valuable insights into rating assessments by incorporating previously unexplored variables pertaining to both service providers and the services itself.

• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.12
• /
• pp.485-496
• /
• 2018

So far, the deep learning, a field of artificial intelligence, has achieved remarkable results in solving problems from unstructured data. However, it is difficult to comprehensively judge situations like humans, and did not reach the level of intelligence that deduced their relations and predicted the next situation. Recently, deep neural networks show that artificial intelligence can possess powerful relational reasoning that is core intellectual ability of human being. In this paper, to analyze and observe the performance of Relation Networks (RN) among the neural networks for relational reasoning, two types of RN-based deep neural network models were constructed and compared with the baseline model. One is a visual question answering RN model using Sort-of-CLEVR and the other is a text-based question answering RN model using bAbI task. In order to maximize the performance of the RN-based model, various performance improvement experiments such as hyper parameters tuning have been proposed and performed. The effectiveness of the proposed performance improvement methods has been verified by applying to the visual QA RN model and the text-based QA RN model, and the new domain model using the dialogue-based LL dataset. As a result of the various experiments, it is found that the initial learning rate is a key factor in determining the performance of the model in both types of RN models. We have observed that the optimal initial learning rate setting found by the proposed random search method can improve the performance of the model up to 99.8%.

• Journal of the Korean Geographical Society
• /
• v.38 no.2
• /
• pp.298-311
• /
• 2003

Building the new administration capital city, one of presidential candidate Moo-Hyun Rho's election pledges, is now listed as a priority national policy agenda of the Participation Government. However, so many people's negative attitudes ranging from cynical skepticism to firm objections against the national policy agenda may threaten its smooth policy actualization. At this juncture, this paper attempts to present persuasive arguments and discuss some critical issues to be considered for building the new administration capital city successfully. The paper begins with taking a look at the current state of hyper-concentration of Seoul agglomeration area and its harmful effects, paints a vision of 'an evenly developed country as a whole' via illustrating the vision from the Swiss case, and reviews the performance of the precedent governments' reform measures for rectifying the hyper-concentration of Seoul agglomeration area. And then, the paper argues for building the new administration capital city as a potent solution to the problem of excessive concentration of activities in Seoul agglomeration area, as well as a driving force to spur the government to realize the Participation Government' enthusiastic vision: 'a decentralized and evenly developed country as a whole' and 'the hub country in the Northeast Asia'. In addition, the paper discusses the location of the new administration capital city in connection with the forthcoming national unification. Lastly, the paper deals with the important issues such as the procedure of people's approval, the population size and legal status of the new administration capital city, the relationship between building the new administration capital city and decentralization reform, etc.

• Advanced Industrial SCIence
• /
• v.3 no.2
• /
• pp.8-16
• /
• 2024

This paper shows the system of drug classification, the goal of this is to foretell the apt drug for the patients based on their demographic and physiological traits. The dataset consists of various attributes like Age, Sex, BP (Blood Pressure), Cholesterol Level, and Na_to_K (Sodium to Potassium ratio), with the objective to determine the kind of drug being given. The models used in this paper are K-Nearest Neighbors (KNN), Logistic Regression and Random Forest. Further to fine-tune hyper parameters using 5-fold cross-validation, GridSearchCV was used and each model was trained and tested on the dataset. To assess the performance of each model both with and without hyper parameter tuning evaluation metrics like accuracy, confusion matrices, and classification reports were used and the accuracy of the models without GridSearchCV was 0.7, 0.875, 0.975 and with GridSearchCV was 0.75, 1.0, 0.975. According to GridSearchCV Logistic Regression is the most suitable model for drug classification among the three-model used followed by the K-Nearest Neighbors. Also, Na_to_K is an essential feature in predicting the outcome.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.40 no.1
• /
• pp.57-64
• /
• 2017

In order to reduce damages to major railroad components, which have the potential to cause interruptions to railroad services and safety accidents and to generate unnecessary maintenance costs, the development of rolling stock maintenance technology is switching from preventive maintenance based on the inspection period to predictive maintenance technology, led by advanced countries. Furthermore, to enhance trust in accordance with the speedup of system and reduce maintenances cost simultaneously, the demand for fault diagnosis and prognostic health management technology is increasing. The objective of this paper is to propose a highly reliable learning model using various machine learning algorithms that can be applied to critical rolling stock components. This paper presents a model for railway rolling stock component fault diagnosis and conducts a mechanical failure diagnosis of motor components by applying the machine learning technique in order to ensure efficient maintenance support along with a data preprocessing plan for component fault diagnosis. This paper first defines a failure diagnosis model for rolling stock components. Function-based algorithms ANFIS and SMO were used as machine learning techniques for generating the failure diagnosis model. Two tree-based algorithms, RadomForest and CART, were also employed. In order to evaluate the performance of the algorithms to be used for diagnosing failures in motors as a critical railroad component, an experiment was carried out on 2 data sets with different classes (includes 6 classes and 3 class levels). According to the results of the experiment, the random forest algorithm, a tree-based machine learning technique, showed the best performance.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.18 no.2
• /
• pp.83-90
• /
• 2010

The windshield wiper consists of 4 parts: a blade, an arm, a linkage and a motor. The wiper blade makes contact with the windshield and is designed to be operated normally at an angle of 30~50 degrees to the front glass. If the contact pressure between the wiper blade and windshield surface is too high, noise and wear of the rubber will result. On the other hand, if the contact pressure is too low, the performance will do badly, since foreign substances such as dust and stains will not be removed well. The pressure and friction of the wiper blade has a great influence on its effectiveness in cleaning the front window. This is due to the contact of the rubber with the window. This paper presents the dynamic analysis method to estimate the performance of the flat type blade of the wiper system. The blade has a nonlinear characteristic since the rubber is an incompressible hyper-elastic and visco-elastic material. Thus, Structural dynamic analysis using a complex contact model for the blade is performed to find the characteristics of the blade. The flexible multi-body dynamic model is verified by the comparison between test and analysis result. Also, the optimization using the central composite design table is performed.

• Journal of KIISE:Computer Systems and Theory
• /
• v.36 no.4
• /
• pp.239-246
• /
• 2009

In this study, we analyze the mean waiting time on the nested open queueing network, where the population within each subnetwork is controlled by a semaphore queue. The queueing network can be transformed into a simpler queueing network in terms of customers waiting time. A major characteristic of this model is that the lower layer flow is halted by the state of higher layer. Since this type of queueing network does not have exact solutions for performance measure, the lower bound and upper bound on the mean waiting time are checked by comparing them with the mean waiting time in the transformed nested queueing network. Simulation estimates are obtained assuming Poisson arrivals and other phase-type arrival process, i.e., Erlang and hyper-exponential distributions. The bounds obtained can be applied to get more close approximation using the suitable approach.